Generative artificial intelligence (AI) is undeniably the biggest tech megatrend in recent memory. And it has created plenty of millionaires and billionaires out of investors who bet big on the right companies at the right time.
That said, not every generative AI stock has been a winner. SoundHound AI (SOUN +1.50%) has struggled to hold Wall Street's attention. And despite having a compelling business strategy in the speech and voice AI niche, shares have fallen by a blistering 41% since the start of 2026.
Let's explore the reasons why the stock is crashing and why it may be on the cusp of a turnaround over the next few years.
Key Data Points
A compelling business model
Large language models (LLMs) like OpenAI's ChatGPT excel at learning information and communicating with the user in a natural, lifelike way. This technology complements voice recognition software, opening up a variety of real-world applications ranging from restaurant drive-thrus to automotive assistants and eventually even humanoid robots. SoundHound has already targeted those first two use cases.
The company offers a software service called Dynamic Drive-Thru, which uses LLMs to help restaurants automate their operations. With food and labor costs rising across the U.S., such systems can be a way for restaurants to maintain their profit margins. Dynamic Drive-Thru is also designed to prioritize efficiency and upselling to boost client sales.
SoundHound has also made a big splash in the market for automotive assistants, which are AI-powered voice-activated software systems that allow drivers to interact with their vehicles hands-free. The company has partnered with European automotive giant Stellantis to install this feature across its DS brand line of vehicles. SoundHound is also working with Hyundai, Honda Motor, and other major global automotive brands.
A widening economic moat
SoundHound has secured multiple high-profile deals with respected brands across different industries. And this suggests that its software could have an economic moat (such as higher quality) that helps set it apart from the alternatives. This is important because SoundHound is not the only company capable of making speech recognition software.
The company also doesn't train or develop the LLMs that power its systems --- instead, it relies on third-party APIs created by AI specialists like Perplexity and Tencent Intelligent Mobility, which help with its automotive assistants. This business model allows SoundHound to focus on making its speech recognition as advanced as possible while other companies provide the most powerful LLMs to integrate with it.
Where will SoundHound be in three years?
Image source: Getty Images.
SoundHound's fourth-quarter earnings were a slam dunk. Revenue soared 59% year over year to $55.1 million, driven by increasing adoption of its automotive assistants in overseas markets like Japan, South Korea, and Italy, where an unnamed sports car manufacturer (possibly Ferrari or Lamborghini) used its software to replace its old provider.
The company has also seen success in other industries like restaurants, where it is partnered with popular U.S. chains like IHOP and Panda Express, and the finance industry, where it is working with BNP Paribas.
SoundHound's diversified revenue streams could add a layer of safety to its business. Furthermore, software-as-a-service (SaaS) contracts tend to be sticky because of high switching costs. So these deals could lay the groundwork for even more growth over the next three years as its clients incorporate voice AI into more aspects of their operations.
The most exciting thing about SoundHound is its bottom line. The company is sprinting toward sustainable profitability with fourth-quarter operating losses shrinking to $42.6 million -- down from a $257.1 million loss in the prior-year period. While shares remain richly valued at a price-to-sales (P/S) multiple of roughly 16, now looks like a great time for investors to start betting on a long-term turnaround as the business scales up.
